About
I am an undergraduate student at the University of Science and Technology of China (USTC).
My research interests lie in video understanding and VLM/LLM post-training. I am especially interested in helping machines discover meaningful structure in long, complex visual sequences.
I am currently open to research collaborations and internship opportunities. Please feel free to get in touch.
Selected Publications
View All →FIS-OT: Feature-Induced Optimal Transport for Unsupervised Action Segmentation
Linxiang Peng, Xinyao Qin, Jinhan Li, Di Yang, Jiangtao Wang
2026 IEEE International Conference on Multimedia and Expo (ICME)
A feature-induced optimal transport framework that couples local representation learning with global structural alignment through cyclic optimization, reducing rigid priors and confirmation bias in unsupervised action segmentation.
News
Our paper FIS-OT has been accepted to IEEE ICME 2026 🎉
Welcome to my homepage!